Many data storage systems include one or more different types of memory architectures and/or implementations. For example, a storage system may be distributed, may be virtual, and/or may provide memory with varied latencies. In many conventional implementations, data is often stored in expensive, low-latency storage devices with capabilities that may exceed the needs of those accessing the data. Similarly, data is often stored in less expensive, higher-latency storage devices that may not suit the needs of those accessing the data. Additionally, data stored in the storage system may be attached to one or more processors such that it is readily accessible, or it may be detached for later use. However, data that is detached may often be stored in low latency memory, sometimes wasting valuable resources, for suboptimal periods of time. As such, finding improved ways to manage and/or place data continues to be a priority.